Flashlight size? What's too big?

[MrNaz]

I want to build a flashlight, but I don't know if my design is too large. I am trying to make a very bright light with lots of run time. I would like to use 6x 18650 batteries arranged in a 3P2S arrangement. The formula to calculate the inner diameter required to fit these batteries is:

Dt = 2r + 2r/cos(30)
= 18 + 18 / cos(30)
= 38.8mm

I am therefore making the inner diameter of the battery tube 40mm. Assuming I make the battery tube walls 2mm thick, that results in an outer diameter of 44mm.

Is a 44mm dia flashlight body too large? I've looked at 17650 batteries, but they sacrifice almost half the capacity for just one mm of diameter saving. Anything thinner than that is just not worth it, and I may as well just use 2S 18650s.

So, 44mm too big? What do other CPFers thing? What is the diameter of a 3D/4D Mag?

 

[DM51]

You might need to add a little to those figures, as protected 18xxx cells are usually slightly over 18 mm diameter. You might consider ~42 mm ID, ~46 mm OD.

But this is a very workable size, IMO. To give you an idea, the Fivemega Megalennium body is roughly that size - in fact it is 47 mm OD at its widest point. I (personally) find it very good to hold, although my hands are a bit larger than average.

 

[MrNaz]

You might need to add a little to those figures, as protected 18xxx cells are usually slightly over 18 mm diameter. You might consider ~42 mm ID, ~46 mm OD.

I have a few 18650s here, and they're all bang on the money at 18mm according both to my eyeballing with a metal ruler and my cheapo micrometer. Also, solving for 40mm means that the cells would have to be over 18.5mm in diameter to not fit.

Given that I plan on shipping the light with batteries that I will source myself, I don't see this as a problem, I would rather a tight, rattle-free design than the ability to accommodate out-of-spec batteries.

This is, however, my first build, so if I'm being naive then PLEASE say so. Don't let me sped $$$ on CNC work if I'm being silly.

But this is a very workable size, IMO. To give you an idea, the Fivemega Megalennium body is roughly that size - in fact it is 47 mm OD at its widest point. I (personally) find it very good to hold, although my hands are a bit larger than average.

This is great to know. I'm not planning on building an EDC, this is going to be a heavy-duty work light, and if the design works out I will be having a batch made in Ti.

 

[DM51]

Obviously I don't know what cells you have, but if you are going to sell these lights, you'll need to allow for other people using different cells that may be slightly larger, especially if yours are unprotected. 18650s are notorious for having slightly different measurements from one brand to another.

People won't necessarily always use the ones you ship with the light, and some people won't want your cells at all - they may prefer to use the ones they already have.

I've just measured an AW protected 18650 and it just nudges the caliper to 18.4mm where the contact strip goes down the side. My caliper is a fairly cheap one, so please don't take that as a definitively accurate measurement.

If I was doing a build, I would definitely err on the side of making the battery compartment slightly too large rather than too small. It's very easy to wrap a piece of paper round a cell if rattling is an issue, but it is not so easy to make the cell slimmer! The last thing people will want is a cell stuck in the tube due to over-heating or because the fit is just too tight.

 

[MrNaz]

I've just measured an AW protected 18650 and it just nudges the caliper to 18.4mm where the contact strip goes down the side. My caliper is a fairly cheap one, so please don't take that as a definitively accurate measurement.

If I was doing a build, I would definitely err on the side of making the battery compartment slightly too large rather than too small. It's very easy to wrap a piece of paper round a cell if rattling is an issue, but it is not so easy to make the cell slimmer! The last thing people will want is a cell stuck in the tube due to over-heating or because the fit is just too tight.

Thanks for that. I'm just wary of making the torch rattle. I don't want to sound contrary, as you are far more experienced than I, however, I just want to make doubly sure that 40mm is not enough. 40mm would allow for 18650 batteries to be up to 18.6mm in diameter. 42mm would allow for batteries up to 19.5mm. Isn't that erring a little too far on the large side? Do 18650s really vary that much in manufacturing?

Also, if I want to use standard AS oring sizes, 42mm ID means a 46mm OD, meaning I would need a 44mm ID oring. In the AS standard sizes, there are 41mm and 46mm orings, but nothing in between. Thus, a 40mm ID/44mm OD tube would be easiest to find orings for.

Does this make sense or am I noobing?

 

[DM51]

Does this make sense or am I noobing?

It all makes very good sense, and no, you are not "noobing"! If you can contemplate making something like this you are already WAY ahead of me in experience, not the other way round.

I'd like to hear an experienced modder/manufacturer give some input here about actual tube diameters - that way you'd hear from someone who really knows what he's talking about. The same goes for O-ring sizes, although I think you'll find there are sources for all conceivable sizes of O-ring. There will be someone here who knows about this too. That's the great thing about CPF - there is always someone around with true expertise, and it's just a case of hoping they will drop by and post in your thread.

I wouldn't get too hung up on rattling - it's not a problem until you put 16 mm dia cells into 18 mm bodies, and even then it's easily solved by wrapping the cell in paper.

 

[rmteo]

So, 44mm too big? What do other CPFers thing? What is the diameter of a 3D/4D Mag?

39.67mm.
http://www.maglite.com/anatomy_spare_D.asp

 

[Benson]

I think 40mm should be fine; the widest point on protected cells is generally over the strip that connects the PCB to the + end, so even if you had some that were a hair over 19.5mm this direction, you can simply load them with that strip oriented away from the wall (and away from the contact point with the other two cells), and they would fit fine.

 

[Al Combs]

Here was an interesting post with a formula you can paste into an Excel spreadsheet. According to it 3*18.6mm diameter 18650's fit in a circle 40.0774 mm diameter. It's trial and error if you're trying to calculate battery size. Here's an exhaustive list from the same thread.

I drilled out my Surefire 6P with an 18.5 mm drill. But I was using unprotected batteries for a direct drive P7. They had an OD of 18.4 mm. If it's any help, my 18650 Solarforce L2 has a tube ID of 19 mm.

 

[baterija]

Do 18650s really vary that much in manufacturing?

From this post old4570 is showing 18.5mm max across 10 different cells. It's a sample size of 1 each but it might help. You might also be overthinking it a little bit worrying about the tolerance of the diameter being so tight to ensure no cell rattle. I assume you are looking at some kind of carrier. As long as the carrier fits tightly enough rattle isn't an issue. Then it's a matter of making sure the carrier can tolerate a wide range of cell. Of course the carrier will separate the cells some increasing the diameter required.

As to the original question, even the slightly larger diameter is manageable in my mind.

 

[lctorana]

Can't fault your maths.
(I used a different formula, starting from an equilateral triangle and the cosine rule, and which resolves to the same result.)

Just a suggestion, since you ask "Is a 44mm dia flashlight body too large?", there is a commercial flashlight with a 3x18650 barrel sold - the UltraFire TH-1300, from one of the dealers we dare not name, or at least not link to.

The brand doesn't enjoy the highest reputation on thse boards for quality or consistency, but that's beside the point for this exercise. It's relatively expensive, but I recommend buying one to feel how a barrel that thick feels in the hand before you proceed...

 

[MrNaz]

Can't fault your maths.
(I used a different formula, starting from an equilateral triangle and the cosine rule, and which resolves to the same result.)

Here are the formulas that I worked out for using to calculating total diameter (Dt) of a circle holding smaller inner circles (Di)

3 Circles:
Dt = Di + (Di / cos(30))

4 Circles:
Dt = Di + Di x sqrt(2)

5 Circles:
Dt = Di + Di / sin(36)

6 Circles:
Dt = 3 x Di

From these I saw that you can generalize the formula to:
Dt = Di + Di / sin(360/2n)

For working out the total diameter of n number of inner circles.

Given the universal usefulness of these, perhaps I should put them in a sticky post...